J. Chem. Eng.
Ostta 1982, 27,
Llteirature Cited Hem,W. J.; Wu, Y. C. J . Phys. Chem. Ref. Data 1972, 7 , 1047. Briiuer, K.; Strehlow, H. Z . Phys. Chem. (Frankfurf am Main) 1958, 77, 346. Akerlijf, 0. J . Am. Chem. Soc. 1930. 5 2 , 2353. AkerlBf, 0.; Turck, H. E. J . Am. Chem. SOC. 1935, 5 7 , 1746. Bates, R. G.; Robinson, R. A. "Chemical Physics of Ionic Solutions"; Wlley: New York, 1966; Chapter 12. Hills, G. J.; Ives, D. J. 0. "Reference Electrodes Theory and Practice"; Academlc Press: New York, 1961; Chapter 3. Bates, R. G. "Determination of pH"; Wlley: New York, 1964; Chapter 9. Sen, 8.; Johnson, D. A.; Roy, R. N. J . Phys. Chem. 1967, 77, 1523.
109
109-113
(9) DIU, A. J.; Itzkowitz, L. M.; Popovych, 0. J . Phys. Chem. 1968, 72, 4580. (10) DIII, A. J.; Popovych, 0. J . Chem. Eng. Data 1969, 7 4 , 240. (11) Berne, D. H.; Popovych, 0. J . Chem. Eng. Data 1972, 77, 178. (12) LaBrocca, P. J.; Phillips, R.; Goldberg, S. S.; Popovych, 0. J . Chem. Eng. Data 1979, 24, 215. (13) Robinson, R. A.; Stokes, R. H. "Electrolyte Solutions"; Butterworth: London, 1955; p 494. (14) Feaklns, D.; Voice, P. J. J . Chem. SOC.,Faraday Trans. 7 1972, 68, 1390. Received for review June 15, 1981. Revised October 30, 1981. Accepted November 20, 1981. This research was supported in part by Grant No. 13138 from the PSC-CUNY Research Award Program of the City University of New York.
Densities and Molar Volumes of the PbCI,-KCI-NaCI PbCI,-KCI-LiCI Ternary Systems
and
Arturo Gutierrer and James M. Toguri' Department of Metallurgy and Materials Science, University of Toronto, Toronto, Canada M5S IA4
Densltles of the molten PbCI,-KCI-NaCI and PbCI,-KCI-LICI systems have been measured by a bottom-balance Archlmedean technlque uslng a twln molybdenum cyllnder denslty bob. From the measured densltles, over a temperature range of 723-1131 K, molar volumes were calculated. The densltles of these melts Increase with Increasing PbCI, content. The ternary molar volumes were found to obey approxlmately a slmple addltlvlty expresston of the blnary molar volumes of PbCI,-KCI and PbC1,-NaCI for the ternary PbCI,-KCI-NaCI system and PbCI,-KCI and PbCI,-LICI for the ternary PbCI,-KCI-LICI system. Introduction The introduction of strict legislation controlling the emissions of sulfur oxides and heavy metals has resulted in a number of innovative metallurgical processes. I n the case of lead production from sulfides, a hydrometallurgical process has been developed by Wong and co-workers in which lead is recovered as lead chloride (1-3). This is then electrolyzed in a fused salt cell to obtain molten lead metal. However, lead chloride alone is not a good electrolyte because of excessive fuming, high melting point, low conductivity, and a high solubility for lead. Marked improvement in these properties is obtained on addiin of alkali chlorides. The ternary systems PbCI,-KCI-LiCI and PbCI,-KCI-NaCI have been suggested as possible electrolytes. Unfortunately, many of the physical properties of these ternary systems are not available. The phase diagrams of the ternary systems clearly indicate the depression of the melting point of lead chloride (501 "C) by the addition of the alkali halides (4). Although the densities of the respective binary systems are known, knowledge of this property within the ternary is not available (5).For this reason, the present study was undertaken to determine the densities of melts in the PbCI,-KCI-LiCI and PbCI,-KCI-NaCI systems.
Experlmental Sectlon Apparatus. The apparatus used to measure the densities of the molten salts is similar to that described previously by Kaiura and Toguri (6). The method is based on Archimede's 0021-9568/82/1727-0109$01.25/0
principle whereby the density of the fluid was determined by measuring the buoyant force of the liquid on twin molybdenum bobs of known volumes. This method differs from the conventional technique in that a balance is placed at the bottom to weigh the molten salt and bob. The apparatus is shown in Figure 1. An alumina reaction tube was placed in a furnace with a constant-temperaturezone of 8 cm at 800 f 3 OC. The reaction tube was sealed at the top by a water-cooled brass end cap with a single gas outlet. A pair of brass plates and a neoprene O-ring seal combined to permit a relatively free, vertical rotational movement of the alumina thermocouple shaft supporting the density bob. Sufficient free play was designed into the seal to allow the thermocouple shaft to act as a limited pendulum. This feature was important in detecting any contact between the wall of the crucible, which contained the fused salt, and the density bob. The bottom of the reaction tube was fiied to a water-cooled brass assembly which also served to seal the Plexiglas balance case housing the Mettler balance. Two argon gas inlets were installed in this apparatus. The one located in the lower part of the Plexiglas case served for flushing this case and for keeping the balance at constant temperature. The other gas inlet located in the brass connection between the alumina reaction tube and the balance case was directed through an alumina tube to a position close to the sample location. Dry, purified argon gas was injected continuously through both inlets, providing a dry, oxygen-free atmosphere throughout the whole system. An alumina sheath, containing a Pt-Pt-13% Rh thermocouple was positioned parallel to the gas inlet tube. Cast onto both of these tubes was a series of alumina radiation baffles. The crucible platform was connected to the Mettler balance by a 10-mm diameter alumina tube which in turn was supported by a brass platform. The heart of the density apparatus was the Mettler PL 1200 electronic toploading balance with a rated precision of f0.005 g and with a readability of 0.01 g. The instrument provided a new weight reading every 0.4 s, which was adequate for these experiments. Typical dimensions of the twin molybdenum density bob are given in Figure 2. The density bob was constructed from two sections of 1.27 cm in diameter, solid molybdenum rod, with a length of 1.5 f 0.002 cm. A chrome1 wire 0.025 cm in dam0 1982 American Chemical Society
110 Journal of Chemical and Engineering Data, Vol. 27, No. 2, 7982
0
296
1 i
292
I
'
I
I
io
C
30
20
4 0
P O S I T I O N OF THE BOB
SELATIVE
(
cm I
Flgure 1. Density-measuringapparatus. BALANCE
READINGS
IN ACTUAL
MEASUREMENT
Figure 3. Density measurements.
___
CRUCIBLE
__-
,iQUID
minimum liquid depth of 4.0 cm was aimed for in all cases. @erafhg prooekne. The following procedure was adopted. The bottom balance supporting the crucible with the salt was kept in a dry atmosphere until the temperature was stabilized. Then weight recordings were obtained for the following sequence of positiins of the calibrated bob: both cylinders of the density bob out, lower cylinder immersed, W,, and both cylinders immersed, W2.A typical balance reading vs. bob position with respect to the crucible is shown in Figure 3. A plateau occurred when the wire connecting the two bobs was at the liquid surface. W , and W, can be defined by the following relations:
LEVEL
W , = V,p UPPER A N D MOLYBDENUW
1
1--15
LOWER
W , = (V,
+ V,
+ surface tension effect on wire (1) + v ) p + surface tension effect on wire (2)
CYLINDERS
4
36 _ _ CRiiClBLE
PLATFORM
SUPPORTING ROD
where V, and V, are the volumes of the lower and upper cylinders, v is the volume of the wire joining both cylinders, and p is the density of the liquid. I f V,, V, and vare known, the density is obtainable without knowledge of the surface tension effect on the wire, since w2
LOWER THERMOCOUPLE
Simensions in m m
Flgure 2. Density bob and crucible.
eter was used to join the bobs. The wire separated the bobs by a distance of 0.8 cm. When both bobs were fully immersed, as shown in Figure 2, the distance separating the thermocouple junction from the liquid surface was at the most 4 mm. Alumina, Mcoanel998 ACN 100, cylindrical crucibles of 9-cm height and 100-mL capacity were used to contain the salt. A
= (V,
- Wl
+ v,+
v)-
v,
-
w2-
Wl
v,+
v
(3)
The linear thermal expansion coefficient of molybdenum is 5 X IO-* cm/(cm deg) (7). Thus, the room-temperature bob volume, V,,, was converted to that at the density-measuring temperature, Vr, according to the equation Vr = V,[1
+ (5 X
10-')ATI3
(4)
Error Analysis. Callbratlon Error. The error introduced during the calibration of the density bobs was due to the readability of the balance. The error was estimated from the average fluctuation between calibration using water and mer-
Journal of Chemlcal and Englneering Data, Vol. 27, No.
2, 1982 111
Table I. Densities (g/cm3) of KCI-NaCI (50-50 mol %) temp, K
this work
ref 8
difference, %
1040 1100
1.548 1.509
1.542 1.509
0.41
0.00
0.2
cury and was found to be f0.35%. The densities of both calibrating fluids have been well established (7) and were not considered as significant error sources. Temperature Error. The difference between the temperature measured by the bob thermocouple and that of the lower thermocouple positioned below the crucible platform corra sponded to f0.3% error in the recorded sample temperature. Balance Error. The balance readability of 0.01 g in combination with the manual recording of the weight changes during immersion of the bob translated to an error of 4~0.25%. Composnlonal Error. The main source of compositional error was due to the volatilization of PbCI,. However, this loss was monitored by the weight loss of the sample, assuming PbC1, to be the only volatile species. The maximum compositional error due to volatilization was f 1%, as determined experimentally. Total Error. The total estimated error in the present study
L
T 025
KClO
1 0L l C I
075
0 50
-t Flgure 4. Pattern used for selection of composition.
is f1.90%. preparadkn. The mixing and the weighing of all salts was carried out in a drybox under an argon atmosphere. The source of the salts used was as follows. Potasskm, ChloMe and Sodium Chloride, Fisher ACS and USP grades were used for potassium chloride and sodium chloride, respectively. These materials were dried for 3 days under vacuum at a temperature of 130 OC and then kept in the drybox. Ltthlum Chloride. 99.8% LiCi obtained from CERAC was used. The lithium chloride which was packed in an argon atmosphere was opened inside the drybox and used without further purification. Lead Chloride. Reagent-grade PbCI, obtained from BDH was used. This material was always handled in a dry atmosphere of the drybox. PreRnlnary Expwhents. The density apparatus was tested by measuring the density of a 50 mol % mixture of NaCI-KCI. The results of this study are shown in Table I. Agreement with the results of Van Artsdalen and Yaffe (8) using the Archima dean method was excellent, the difference being 0.41% at 1040 K.
Results The compositions of the ternary systems are expressed by the parameters y and t ( 9 ) ,as shown in Figure 4, where Y = 1 - Xm,* and
36 450
500
550
600
650
180
TE ' , Y i
Figure 5. Typical results of density change with temperature. where M = Li or Na. In these equations, Xi is the mole fraction of component i. Eleven compositions in the PbCI,-KCI-LiCI system were studied. Typical results are shown in Figure 5 for a composition corresponding to y = 0.4 and t = 0.5. Using a least-squares technique, we fitted the density results to a straight-line equation of the form p=A+BT
(7)
where A and B are constants and Tis the absdute temperature (K). The tabulation of the constants A and 8 , the temperature range of the measurements, and density values predicted for certain temperatures are shown for each experimental composition in the PbC1,-KCI-LiCi and PbCI,-KCi-NaCI systems in
Table 11. Densities for the PbCI,-KCI-LiCI Ternary System density, g/cm3
compn, mol o/c PbCl,
KC1
LiCl
A
80 80 80 60 60 60 40 40 40 20 20
5 10 15 10 20 30 45 30 15 20 40
15 10 5 30 20 10 15 30 45 60 40
5.578 5.296 5.514 4.957 5.015 4.814 4.060 4.066 4.195 3.471 2.932
104~ -14.0 -11.0 -15.0 -12.2 -14.0 -12.5 -10.6 -9.8 -10.4 -10.9 -7.5
temp, "C
673 K
723 K
113 K
823 K
873 K
923 K
973 K
1073 K
527-675 475-622 464-547 527-675 440-716 45 2-6 83 445-126 502-780 414-7 19 60 7-7 35 485-712
4.636 4.556 4.564 4.136 4.073 3.972 3.346 3.406 3.501 2.731 2.427
4.566 4.501 4.984 4.075 4.003 3.909 3.294 3.351 3.450 2.683 2.389
4.496 4.446 4.414 4.014 3.933 3.847 3.241 3.308 3.398 2.628 2.352
4.426 4.391 4.339 3.953 3.863 3.184 3.188 3.260 3.347 2.514 2.315
4.356 4.336 4.264 3.892 3.793 3.722 3.135 3.210 3.296 2.519 2.277
4.290 4.281 4.189 3.831 3.723 3.659 3.081 3.161 3.244 2.465 2.240
4.216 4.226 4.114 3.770 3.653 3.597 3.028 3.112 3.193 2.411 2.202
4.076 4.116 3.964 3.648 3.513 3.472 2.922 3.015 3.089 2.302 2.127
112 Journal of Chemical and Engineering Data, Vol. 27, No. 2, 1982
Table 111. Densities for the PbC1,-KCI-NaCI Ternary System
compn, mol R
density, g/cm3
PbCI,
KC1
NaCl
A
20 20 20 42.85 42.85 40 60
20 40 60 14.30 42.85 30 20
60 40 20 42.85 14.30 30 20
3.219 3.105 3.219 4.235 4.098 3.969 4.920
60
104B -8.2 -7.4 -10.1 -10.4 -10.5 -9.4 -13.7
+I 0
temp,"C
673 K
723 K
773 K
823 K
873 K
923 K
973 K
1073 K
721-852 702-857 717-845 647-795 617-795 582-747 487-785
2,667 2.607 2.539 3.535 3.391 3.336 3.997
2.626 2.570 2.488 3.483 3.338 3.289 3.929
2.585 2.533 2.438 3.431 3.286 3.242 3.860
2.544 2.496 2.387 3.379 3.233 3.195 3.792
2.503 2.439 2.337 3.327 3.181 3.148 3.723
2.462 2.422 2.286 3.275 3.128 3.101 3.655
2.421 2.385 2.236 3.223 3.076 3.054 3.586
2.339 2.311 2.135 3.119 2.971 2.960 3.449
Table IV. Molar Volumes in the PbC1,-KCI-LiCI System :1:
V = C + DT," cm /mol
~
compn, mol %
t. 0
PbCl,
KCl
LiCl
C
D
80 80 80 60 60 60 40 40 40 20 20
5 10 15 10 20 30 15 30 45 20 40
15 10 5 30 20 10 45 30 15 60 40
38.500 42.346 39.040 34.574 34.614 37.277 31.602 33.466 34.423 24.306 32.538
0.0171 0.0134 0.0187 0.0143 0.0179 0.0169 0.0130 0.0140 0.0159 0.0159 0.0142
"TinK. Table V. Molar Volumes in the PbC1,-KCI-NaCI System V = C -k DT," cm3/mol
compn, mol "%
L Tables I I and I I I , respectively.
PbCl,
KCl
NaCl
c
D
60 42.85 40 42.85 20 20 20
20 14.30 30 42.85 20 40 60
20 42.85 30 14.30 60 40 20
35.876 34.225 36.186 36.386 30.656 32.987 30.633
0.0185 0.0142 0.0135 0.0158 0.0132 0.0129 0.0199
T i n K.
Discussion From the density data, the molar volumes of the ternary mixtwes were evaluated. The molar volume of the mixture, V, is defined by I
1.0and show an ideal behavior. Atomistically, the exchange of alkali cations K+ for Li+ or K+ for Na' results in a linear change in molar volume, at a constant Pb2+content. This can be expressed by a simple additivity relationship
CXIMI
I v= -
P
(9) (8)
where M , = molecular weight of component i. Typical calculated molar volumes for the ternary mixtures PbCI,-KCI-LiCI are shown in Figures 6 and 7 for temperatures of 773 and 973 K, respectively. In Figure 6, the mdar volumes are shown as a function of y at constant t . The curves for t = 0 and t = 1 correspondto the binary systems PbCI,-KCI and PbCI,-LiCI, respectively, and were evaluated from the data reported by Boardman et al. (70) and Shevlyakova and Lantratov ( 7 7 ) . In the case of the binary PbCI,-alkali chloride systems, increasing positive deviatlons are observed with an increase in the ionic radius of the alkali cation. This positive deviation can be attributed to the stabilization of the anionic complex PbCI,- as the alkali cation radius is increased from Li+ to CS+. I n both the PbCI,-KCI-LiCI and PbCI,-KCI-NaCI ternary systems, a linear change in molar volume with t was observed at constant y, within experimental error, as shown in Figure 7 for the case of the PbCI,-KCI-LiCI system at a temperature of 973 K. The molar volumes in the KCI-LiCI binary system reported by Van Artsdalen and Yaffe (8) are represented by y =
where
Vi,wcv) = molar volume in the ternary system at y
= molar volume in the 1-2 binary system at y V I w ) = molar volume in the 1-3 binary system at y A similar equation was derived by Carmichael and Flengas (72). On the basis of this relationship, isovolumetric lines can be constructed over the entire liquid region of the ternary diagram at constant temperature. Typical examples are shown in Figures 8 and 9 for the PbCI,-KCI-LiCI and PbCI,-KCI-NaCI systems at a temperature of 773 and 973 K, respectively. The experimentally obtained molar volumes are shown by solid circles. For the binary system NaCI-PbCI,, the data reported by Bloom et al. (73)were adopted. A summary of the calculated molar volumes is given in Tables I V and V. Although the density data indicate that these systems are relatively simple in their physical behavior, this assumption may not be completely valid. As indicated by the emf studies of Hagemark et al. ( 1 4 ) , negative deviations in GEwI, were observed in the PbC12-KCI-NaCI system as the PbCI,-KCI binary
Journal of Chemical and Engineering Data, Vol. 27, No. 2, 1982 113
Flgure 9. Isovolumetric lines in the PbCI,-KCI-NaCI
system at 973
K.
(2) The density data were used to calculate the molar voiumes. The molar volumes were found to be additive at constant values of y, and their behavior could be described by the additivity rule with respect to t : 0
0.2 5
0.50
0.7 5
V123(y) =
1.0
tFigure 7. Molar volumes as a function of t at constant y and at 973 K.
(1 -
W13(y)
+ tV12W)
where the ternary system is designated by the subscripts 123, and the two binary systems, each containing PbCi,, are denoted by the subscripts 12 and 13.
Acknowledgment We are grateful to Dr. J. E. Dutrizac for many helpful discussions.
Llterature Clted (1) Haver, F. P.; Elges, C. E.; Wong, M. M. Rep. Invest.-US., (2) (3) (4)
(5) (6) (7) (8) (9)
Figure 8. Isovolumetric lines in the PbCI,-KCI-LiCI
system at 773 K.
was approached, with the maximum deviation occurring near the 33 mol % PbCI, composition.
Concluslons (1) A twin density bob bottom balance technique was used to determine the densities of the ternary systems PbCI,-KCINaCl and Mi,-KCI-Lici over the temperature range 723-1 130 K. The density values were fitted to a linear expression of temperature.
(10)
Bur. Mines 1978. R I 8166. Murphy, J. E.; Haver, F. P.; Wong. M. M. Rep. Invest.-US., Bur. Mines 1974, RI 79 13. Wong, M. M.; Haver, F. P. “Symposium on Molten Salt Eiectrolysls in Metal Production”; Institute of Mining and Metallurgy: London, 1977. Levin, E. M.; Robbins, C. R.; McMwdle. H. F. “Phase Dlagram for Ceramists”; American Ceramic Society, Inc.: Columbus. OH, 1964; Supplement 1969. Janz, G. J.; Tomkins, R. P. T.; Ailen, C. 6.; Downey, J. R.; Gardner, G. L.; Krebs, U.; Singer, S. K. J. Phys. Chem. Ref. Data 1975, 4 , 871. Kaiura, G. H.; Togurl, J. M. Can. Metall. P. 1979, 18, 155. Weast, R. C., Ed. “Handbook of Chemistry and Physics”, 56th ed.; CRC Press: Cleveland, OH, 1975. Van Artsdaien, E. R.; Yaffe. I. S. J. Phys. Chem. 1955, 59. 118. Wagner, C. “Thermodynamics of Alloys”; Addison-Wesley: Reading, MA; 1952; p 15. Boardman, N. F.; Dorman, F. H.; Heymann, E. J . Phys. Chem. 1949, 53. - - , 375. - . -.
(11) Shevlyakova, T. N.; Lantratov, M. F. T r . Vses. Soveshch. Fir. Khim. Rasplevl. Solel. 2nd. 1963 1965, 8%. (12) Carmichael, N. R.; Fiengae. S. N. J . Electrochem. Soc. 1979, 126, 2098. (13) Bloom, H.; Boyd. P. W. 0.; Laver, J. L.; Wong. J. Aust. J . Chem. 1988, 19, 1591. (14) Hagemark, K.; Hengstenberg, D.; Blander, M. J. Chem. €ng. Data 1972, 17, 216. Received for review December 1, 1980. Revised manuscript recehred June 8, 1981. Accepted November 3, 1981. The research described in this paper was carried out under contract and with the flnancial support of the Canada Centre for Mineral and Energy Technology, Energy, Mines and Resources, Canada Contract No. OS0 79-00300, Supply and Services, Canada.
